Discussion:
WEAKNESSES:
Throughout our study, we found a few
weaknesses in our study that may have influenced the data we
collected. First, we experienced nonresponse bias. We
sent out 150 samples and received only 123 back. To reduce the
influence of this bias we could have distributed even more samples,
making the sample size larger. This bias could also have been
reduced by waiting longer for surveys to be turned in. Due to
time restraints were forced to cut our sample off at 123 surveys
received. We also dealt with response bias within our sample.
A few of our surveys were filled out in a manner that disqualified
them from our sample. There are not many ways other than
increasing sample size to reduce the effects of this response bias.
There was no measurement bias introduced in this study.
Lastly, we could have strengthened our study by stratifying based on
grade level.
EXTRAPOLATION:
Our sample was large enough to
extrapolate to the population at North Olmsted High School. We
could not extrapolate our study to other high schools or all high
schools across the nation because of our limited sample size and
also the variation of phone preference across different locations.
In order to extrapolate our sample further we would need to include
high schools from multiple different locations in our sample.
Therefore, we required a larger sample in order to be able to apply
the results of the study to a larger population.
SUGGESTIONS FOR FURTHER STUDIES:
For further studies, we could have
potentially determined if a relationship between phone and gender
appeared after grade level was stratified. That is, does a
relationship exist in freshman, sophomores, etc. Additionally, due
to the large number of types of Androids, we could potentially split
that category into multiple to complete a more thorough analysis of
the phone analysis.